At yesterday’s Web Monday, hosted by Metalab, I attended an interesting presentation about microformats (given by Eric). Among other things the open microformat standard includes a “geo” element to define geographic locations.
As Alex yesterday mentioned, the documents, as well as the documents referred to, lack of declaring a reference system. It’s quite obvious, and has been already discussed in the geo microformat BOF, that the coordinates enclosed by the “geo” tag are based on WGS84.
How come that WGS84 didn’t make it into the document then? I think an “open standard specification” must not allow any room for misinterpretations and therefore should include the declaration of the applied geographic reference system in some section.
…it works in Google Maps, so it seems to be right…
Ouch.
Why is defining a reference system important?
Once you leave Google Maps behind and try to visualize your geographic information in other mapping applications, such as national mapping services based on national reference systems others than WGS84 for instance, you should tell those applications about the framework you used to identify your locations. Otherwise you’ll run into troubles regarding the accuracy of your locations.
Austria is a small country but the national reference system includes 3 different prime meridians. In order to achieve satisfying mapping results you’ll always choose the closest meridian to your mapping area. Without the information which meridian was used for data collection for instance, you’ll end up having at least 3 options where your locations could be.
No serious web developer would ever write an HTML-document without a valid Document Type Declaration. It’s quite similar in geodesy with geographic data and reference systems. Would be great to see neogeographers fulfill minimum geodesy requirements. Geography is for some time around now and things like reference systems have turned out to be very useful.